Vol. 64, No 2/2017 227–231 https://doi.org/10.18388/abp.2016_1299 Regular paper The Gas6 gene rs8191974 and Ap3s2 gene rs2028299 are associated with type 2 diabetes in the northern Chinese Han population Elena V. Kazakova#, Tianwei Zghuang#, Tingting Li, Qingxiao Fang, Jun Han and Hong Qiao* 1The Fifth Endocrine Department, the Second Affiliated Hospital, Harbin Medical University, Harbin, Heilongjiang, China Previous studies in other countries have shown that INTRODUCTION single nucleotide polymorphisms (SNPs) in the growth arrest-specific gene 6 (Gas6; rs8191974) and adapt- Diabetes mellitus affects more than 300 million indi- er-related protein complex 3 subunit sigma-2 (Ap3s2; viduals worldwide, with increasing prevalence particular- rs2028299) were associated with an increasedrisk for ly in the developing countries (Whiting et al., 2011). In type 2 diabetes mellitus (T2DM). However, the associ- fact, the prevalence of type 2 diabetes mellitus (T2DM) ation of these loci with T2DM has not been examined in China is among the highest in the world. The com- in Chinese populations. We performed a replication bination of insulin resistance in peripheral tissues and study to investigate the association of these suscep- impaired insulin secretion from pancreatic β-cells is be- tibility loci with T2DM in the Chinese population. lieved to contribute to the development and progression We genotyped 1968 Chinese participants (996 with of T2DM. Both, the genetic and environmental factors T2DM and 972controls) for rs8191974 in Gas6 and confer susceptibility to T2DM. In recent years, studies rs2028299 near Ap3s2, and examined their associa- of gene polymorphisms have helped identify a number tion with T2DM using a logistic regression analysis. of T2DM-susceptibility genes. We also analyzed the correlation of genotypes and The growth arrest-specific gene 6 (Gas6) is located clinical phenotypes. The distribution of the T allele on human chromosome 13q34. Plasma Gas6 is a vita- of SNP rs8191974 in the Gas6 gene was significant- min K-dependent protein very similar to plasma an- ly different between T2DM cases and controls when ticoagulant protein S and is comprised of the Gla do- compared with the C allele (P<0.05, OR: 0.80，95% main (γ-carboxylated glutamicacid-rich region) and four CI: 0.69–0.94). The occurrence of the CT genotype and EGF-likedomains (Bellido-Martin et al., 2008). Gas6 the dominant model was also significantly less fre- protein interacts with receptor tyrosine kinases of the quent in the T2DM cases vs. controls when compared Tyro-3, Axl, Mer(TAM) family (Hafizi et al., 2006). The with the CC genotype (CT vs. CC: P<0.05, OR: 0.75, Gas6/TAM system has a number of diverse functions, 95% CI:0.62–0.90; TT+CT vs. CC: P<0.05, OR:0.75, 95% including the regulation of cell survival and proliferation, CI:0.63–0.90). In SNP rs2028299, the allele C showed cell adhesion and migration, and others (Godowski et al., no statistically significant differencein distribution 1995; Nagata et al., 1996; Bellosta et al., 1997; Collett et between the control and T2DM groups when com- al., 2003;). In recent years, several studies have explored pared with allele A. However, in male populations, the the relationship between Gas6 polymorphisms and dominant model was statistically more frequent when T2DM. Chien-Hsing Lee in Taiwan found that the Gas6 compared with genotype AA (CC+CA vs. AA: P<0.05, gene single nucleotide polymorphism (SNP) rs8191974 OR:1.29, 95% CI:1.02–1.64), and in obesity-stratified was associated with T2DM in the local population (Lee analysis, we also observed a significant difference in et al., 2012). the distribution of the dominant model between the The adapter-related protein complex 3 subunit sig- T2DM cases and controls in subjects with BMI≥24 kg/ ma-2 gene (Ap3s2) is located on human chromosome m2 and BMI<28kg/m2 (CC+CA vs. AA: P<0.05, OR: 6.33, 15q26,and was first foundin muscle tissue of rats. Ap3s2 95% CI:4.17–9.61). In conclusion, our study shows that protein is composed of 193 amino acids and its physio- SNPsrs8191974 and rs2028299 are significantly associ- logical function involves protein transport. Two studies ated with T2DM in the Chinese population. have reported that Ap3s2 gene SNP rs2028299 was as- sociated with T2DM. A genome-wide association study Key words: Type 2 diabetes mellitus; Gas6 gene; Ap3s2 gene; single (GWAS) in people of South Asian ancestry identified nucleotide polymorphisms; epidermal growth factor (EGF)-like rs2028299 as asusceptibility locus for T2DM (Kooner et Received: 09 October, 2016; revised: 26 October, 2016; accepted: al., 2011). A replication study investigating the associa- 05 Januray, 2017; available on-line: 11 April, 2017 tion of SNP rs2028299 with T2DM in the Japanese pop- *e-mail: [email protected] ulation (Fukuda et al., 2012). Found that SNP rs2028299 #These authors contributed equally to this work. had a significant effect on the susceptibility of Japanese Abbreviations: BMI, body mass index; HDL, high-density lipopro- people toT2DM (Fukuda et al., 2012). tein cholesterol; LDL, low-density lipoprotein cholesterol; HOMA-β, To our knowledge, the role of the Gas6 and Ap3s2 homeostasis model assessment of beta-cell function; HOMA-IR, ho- meostasis model assessment of insulin resistance genes inT2DM in the Han population has not been pre- viously assessed. Given the apparent association between SNP rs8191974 and SNP rs2028299 with T2DM, we further investigated whether SNP rs8191974 and SNP rs2028299 are related to T2DM in Chinese people. 228 E. V. Kazakova and others 2017 blood glucose, homeostasis model assessment of insulin Table 1. Clinical characteristics in patients and controls. resistance (HOMA-IR) and HOMA of beta-cell function NGT T2DM P value (HOMA-β). We used the case control study methods to Sex (M/F) 568/404 612/384 0.1731 analyze the association between Gas6 gene rs8191974, Ap3s2 gene rs2028299 and T2DM. The case group in- Age (years) 42.87±11.67 46.11±12.55 <0.0001 cluded 1000 patients with T2DM, and the control group Height 1.69±0.08 1.68±0.08 0.0801 included 1000 healthy human subjects. Inclusion criteria for the control subjects were as follows: 1) Hemoglobin Weight 66.55±12.37 73.13±13.45 <0.0001 A1c level < 6.0%; 2) fasting plasma glucose level < 5.1 BMI 23.31±3.34 25.78±3.58 <0.0001 mmol/L; 3) no history of diabetes; 4) no family history Waistline 81.25±10.84 93.49±10.44 <0.0001 of T2DM; 5) not taking drugs that affect lipid and car- bohydrate exchange; and 6) no systemic diseases. Exclu- Hipline 95.77±7.2 99.48±7.44 <0.0001 sion criteria for this study were as follows: 1) renal and Waist-to-hip ratio 0.85±0.07 0.94±0.06 <0.0001 hepatic failure; 2) cardiovascular disease; 3) acute diabetic complications; 4) malignant tumors, severe injury, infec- Systolic 121.24±15.08 130.17±17.5 <0.0001 tions, or other endocrine diseases; or 5) other types of Diastolic 79.21±9.64 84.64±11.17 <0.0001 diabetes. Fasting glucose 4.82±0.33 10.03±3.41 <0.0001 DNA extraction and genotype analysis. DNA was obtained from blood samples using the TIANGEN TC 4.88±1.05 5±1.29 0.0289 whole blood DNA extraction kit and following the man- TG 1.42±0.96 2.38±2.25 <0.0001 ufacturer’s instructions. The quality of isolated genomic DNA waschecked using agarose gel electrophoresis and HDL 1.48±0.37 1.21±0.32 <0.0001 quantitation wasdetermined using spectrophotometry. LDL 1.92±0.87 2.91±1.26 <0.0001 The Gas6 gene SNP rs8191974 and Ap3s2 gene SNP HOMA-IR 1.51±1.15 4.81±3.91 <0.0001 rs2028299 genotyping was performed using commercial SNPscan Genotyping assays from Applied Biosystems HOMA-β 104.78±74.62 38.42±37.51 <0.0001 Inc. Statistical methods. Descriptive results of continu- ous variables areexpressed as means ±standard error of MATERIALS AND METHODS the mean. Before statistical analysis was performed, nor- mal distribution and homogeneity ofvariables was evalu- Study population. All subjects were recruited from ated by using the Levene’s test, and if necessary, a base outpatient clinics of the Second Affiliated Hospital, logarithmic transformation was applied. The x2-test was Harbin Medical University, Heilongjiang, China. Basic used to determine the genotype distributions for Hardy– parameters were collected from each subject, including Weinberg equilibrium and to compare the observedal- name, age, gender, height, weight, waist circumference, lele and genotype frequencies in patients with those in hip circumference, blood pressure, plasma lipids, fasting control subjects. Statistical significance was considered to be P < 0.05. Logistic regression Table 2. Correlation analysis of Gas6 rs8191974 and Ap3s2 rs2028299 alleles with T2DM. analyses were performed with a a SPSS (Chicago, Illinois, USA; Gene Allele controls(972) T2DM(996) P OR (95%CI) version 13.0) and the odds ra- A 1573(80.92%) 1566(78.61%) tio (OR) and 95% confidence Ap3s2 0.0725 1.1534(0.9870-1.3478) C 371(19.08%) 426(21.39%) interval (CI) were calculated for each variant genotype com- C 1465(75.36%) 1577(79.17%) Gas6 0.0044 0.8049(0.6931-0.9364) pared to the homozygous wild- T 479(24.64%) 415(20.83%) type genotype. Abbreviations: CI, confidence interval: OR, odds ratio.a Adjusted for age, gender and BMI. RESULTS Table 3. Correlation analysis of Gas6 rs8191974 and Ap3s2 rs2028299 genotypes with T2DM. Subject Characteristics Gene Genotype Controls T2DM Pa ORa(95%CI) There were four cases in the AA 637 (65.53%) 613 (61.55%) 1 T2DM group and 28 cases in the control group without gen- CA 299 (30.76%) 340 (34.14%) 0.1809 1.18 (0.98-1.43) otypes, so the final study in- Ap3s2 CC 36 (3.7%) 43 (4.32%) 1.24 (0.79-1.96) cluded 996 T2DM patientsand AA 637 (65.53%) 613 (61.55%) 972 control subjects.